extracting specific line instead of printing whole line using awk/sed

[ryand85]

Hi all,

I understand how to find a specific pattern but how do you "print" that specific pattern your looking for instead printing the whole line that pattern is on. In my case, I have to extract websites address that have a similiar pattern of:

"http://www.address.com"

- all of the addresses in the file start with quotes, followed by "http://" and end with double quotes. So thats the thing i need to match to get those patterns out. But how do I only get the output to print out what i'm looking for and not the whole line?

- A sample of the type of file im dealing with is below:

asdasdsad9999123123===080"http://www.linux.com"kajnsknd123123
ijiasdkkjnas"http://www.ign.com/reviews"nas,mdn123123as

etc....

Any suggestions or help would be greatly appreciated,

Thanks

[ryand85]

Oh yeah i used the following sed command, to extract the pattern "http://"
Hope that helps so you get an idea of what i'm trying to do.

sed -n '/\"http:\/\/.*"/ p' bookmarks.html

[vsemaska]

If you don't mind using a bash script instead here's one. It's crude but it works.

#!/bin/bash

typeset -i _pos
while read _lne
    do
        _pos="`echo "$_lne" | awk -F"^" '{print match($1,"http://")}'`"
        _tmp="`echo "$_lne^$_pos" | awk -F"^" '{print substr($1,$2)}'`"
        _tmp2='"'
        _pos="`echo "$_tmp^$_tmp2" | awk -F"^" '{print match($1,$2)}'`"
        _pos=_pos-1
        _tmp2="`echo "$_tmp^$_pos" | awk -F"^" '{print substr($1,1,$2)}'`"
        echo "$_tmp2"
    done < bookmarks.html

[ghostdog74]

awk 'BEGIN{FS="\""}
{
  for(i=1;i<=NF;i++){
   if( $i ~ /http/ ) print $i
  }
}
' file

output:

# ./test.sh
http://www.linux.com
http://www.ign.com/reviews

[ghostdog74]

It's crude but it works.

awk is a general purpose language readily available in *nix that performs the job of grep,sed,cut,wc to name a few. Learning how to use awk to your advantage will certainly benefit you in future

[ryand85]

thanks for the help!! greatly appreciated

[vsemaska]

awk is a general purpose language readily available in *nix that performs the job of grep,sed,cut,wc to name a few. Learning how to use awk to your advantage will certainly benefit you in future

I manage over 30 servers that are either Red Hat Linux or (Tru64) UNIX. Some of those Tru64 UNIX servers are in one of two TruCluster UNIX clusters which are very complicated. I have to know bash, ksh, csh, perl, plus a couple of other scripting languages plus numerous software products used on the servers. I simply don't have the time or desired to learn another cryptic scripting language.

[jlarvie]

Just to note you could use something like

perl -e 'while(<>) {s#.+?(https?://.+?)".+#$1#;print;}' <file>

[burschik]

grep can do that:

grep -o '"http://[^"]\+"'